|
|
|
|
| Author |
|
|
Network Native
Join Date: Jan 2007
Location: SoCal
Posts: 10,349
|
Pondering some exhaust manifold ideas
Yesterday I was thinking about the exhaust manifold on the 928, the cast iron OB, the stepped log on the 32v, and full length headers. The 1 3 7 2 6 5 4 8 firing order makes things kinda odd, cylinders that are desirable to merge aren't all on the same bank, so plumbing 8 pipes gets messy fast.
What I was wondering is if pairs of cylinders on each bank might be joined that have no time overlap in exhaust flow, like 1+2 and 3+4, and 5+7 and 6+8, aiming for a near eq length at a 4/1 merge pre cat. Without overlap in time, seems to me the pipes can be near or the same size as the primary to the exhaust port, and with just two per side clearance issues should be less. I'm thinking either a sleeve or flange connection on each side for the two pipes so in and out isn't insanity, and I'm not sure what might work best for the 4/1 merge, but anybody have any idea if the basic concept should work? |
||
08-10-2008, 02:22 PM
|
|
|
Network Native
Join Date: Jan 2007
Location: SoCal
Posts: 10,349
|
I can't find images of Tom's exhaust (link from Dr Ott's site is expired), but I remember it was "busy" and I think he brought down all 8 primaries and merged pairs. I think all those bends have a big hit on flow, and that something simple might be a good trade off, less merge gains, but better flow and cheaper to make.
|
||
|
08-10-2008, 03:01 PM
|
|
|
Registered
Join Date: Jan 2007
Posts: 11,758
|
I presume this is what the X pipe does, only further downstream.
I have seen pictures of the exhaust you describe. Real spaggetti. |
||
|
08-10-2008, 03:10 PM
|
|
|
Registered
Join Date: Aug 2006
Location: Cincinnati, Ohio
Posts: 1,019
|
The key is to use the pulses to create a draw for scavenging on a NA car. Check out my manifold pictures...you can see how I keep the pulses seperated between 5 and 6 on the turbo car...those are the two which fire in sequence. With a turbo car there is no scavenging, just go big after the turbo and get it out!
__________________
Kuhn Performance Technologies, LLC Big Gun: 1988 928S4 Twin Turbo, 5-SPD/LSD 572 RWHP, 579 RW ft-lbs, 12 psig manifold pressure. Stock Internals, 93 octane. Little Gun: 1981 928 Competition Package Twin Turbo, 375 RWHP, 415 RW ft-lbs, 10psig manifold pressure. Nikasil Block, JE2618 Pistons, 93 octane. |
||
|
08-10-2008, 04:40 PM
|
|
|
Network Native
Join Date: Jan 2007
Location: SoCal
Posts: 10,349
|
I think I know just enough to be dangerous.
 As I understand it three factors are; basic flow resistance from size of the pipe and bends, tuning from the length of the primary tube from valve to collector, and scavenging with merges. I think for the merge to be effective you need flow from another cylinder to create a low pressure pulse at the merge which I think travels back up the pipe at the speed of sound and if everything is designed right reaches the valve when its open. With all 8 pulses going through the same merge, if the pipe length is good it should work for all cylinders. BTW I picked the pairs to join near the heads on each bank to keep each exhaust pulse as far in time from the other as possible, no scavenging to keep all the cylinders in balance. I don't really understand why on the factory manifold that flow overlaps on 1, 3, and 4, but not on 2, which would make me think 2 won't flow as well as the others. |
||
|
08-10-2008, 10:33 PM
|
|
|
Heavy Metal Relocator
|
180 degree, equal length primaries......
very similar to what's used in NASCAR. I, too, have seen the pictures. Nice plumbing work, but could be very, very expensive unless produced in mass. (remember, a set of MSDS coated headers will set you back $700-800, 928 Int'l headers are ???, and set of Devek are ???, if you can find some) --Russ
__________________
Absence of Evidence, is not Evidence of Absence. Bill Maher 8/4/09--- "I'll show you Obama's birth certificate, when you show me Sarah Palin's high school diploma." |
||
|
08-11-2008, 11:23 AM
|
|
|
